Prototyping describes the process of creating a first-version exhibit, then testing it out with visitors, and redesigning. Often, we iterate this several times, depending on monetary and time budgets. It’s usually a fruitful way to find out not only what buttons confuse people, but also what they enjoy playing with and what great ideas totally bomb with users.
The problem with prototyping, as with many data collection processes, is that you have to ask the right questions to get useful answers. We are currently re-developing an interactive about how scientists use ocean data to make predictions about salmon populations for future harvests. The first round surveys revealed some areas of content confusion and some areas of usability confusion. Usability confusion is easy to re-work usually, but content confusion is harder to resolve, especially if your survey questions were confusing to the visitors.
This was unfortunately the case with the survey I made up, despite a few rounds of re-working it with colleagues. The survey had multiple-choice questions which were fairly straightforward, but it was the open-ended questions that tripped people up, making the results a bit harder to interpret and know what to do with. The moral of the story? Prototype (a.k.a. pilot) your survey, too!